Академический Документы
Профессиональный Документы
Культура Документы
6-1
Sound System Design Reference Manual
Acoustical Feedback and Potential and allow another 3 to 4 dB of system gain. Another
System Gain possibility is to place the listener in the direct field of
the loudspeaker, allowing a further increase in
Just as in the outdoor case studied earlier, system gain. If the level of the reverberant field is
if we have a microphone/amplifier/loudspeaker lower in the performing area than in the listening
combination in the same room and gradually turn up area, additional system gain also results. This
the gain of the amplifier to a point approaching situation is described by the Boners as a room
sustained feedback, the electrical frequency constant in the microphone area different from that in
response of the system changes with the gain the seating area. Similar results may be noted in
setting. The effect results from an acoustic feedback rooms having large floor areas, relatively low
path between the loudspeaker and the microphone. ceilings, and substantial sound absorption. In such
As a person talks into the microphone, the rooms, as we have seen, sound from a point source
microphone hears not only the direct sound from the tends to dwindle off beyond DC at a rate of 2 or 3 dB
talker, but the reverberant field produced by the for each doubling of distance rather than remaining
loudspeaker as well. constant in level.
The purpose of using high-quality loudspeakers Still another way to increase gain is to
and microphones having smooth response electrically suppress the positive feedback
characteristics, and sound system equalization (apart frequencies individually with very narrow bandwidth
from achieving the desired tonal response) is to filters. If one could channel all energy into the
smooth out all of the potential feedback points so negative feedback frequencies, the potential system
that they are evenly distributed across the audible gain would theoretically become infinite! Unfortunately,
frequency range. When this has been done, there the acoustic feedback path is not stable enough to
should be as many negative feedback points as permit this degree of narrow-band equalization.
positive feedback points, and the positive feedback In all other situations, a gain setting is reached
points should all reach the level of instability at about at which sustained oscillation occurs. By definition,
the same system gain. maximum system gain is reached just below this
We might expect this to average out in such a point. However, the system cannot be operated
way that the level produced by the loudspeaker satisfactorily at a point just below oscillation because
reaching the microphone can never be greater than of its unpleasant comb-filter response and the
that produced by the talker without causing sustained prolonged ringing caused by positive feedback
oscillation. In other words, we assume that the extra peaks. To get back to reasonably flat electrical
gain supplied by all the positive feedback spikes is response and freedom from audible ringing, it usually
just balanced out by the loss caused by all the is recommended that a properly equalized system be
negative feedback dips. operated about 6 dB below its maximum gain point.
If the Boner criteria for optimum system Even an elaborately tuned system using narrow-
geometry are followed, the microphone will be close band filters can seldom be operated at gains greater
to the talker so that it hears mostly direct sound from than 3 dB below sustained oscillation.
the talker. It will be far enough from the loudspeaker
to be well into the reverberant field of the
loudspeaker, so that direct sound from the Sound Field Calculations for a Small Room
loudspeaker is not an appreciable factor in triggering
system feedback. Assuming that listeners are also in Consider the room shown in Figure 6-2. This is
the reverberant field of the loudspeaker, it follows a typical small meeting room or classroom having a
that the sound level in the listening area with the volume less than 80 m3. The average absorption
system turned on cannot be greater than that of the coefficient a is 0.2. Total surface area is 111 m2. The
unaided talker at the microphone position with the room constant, therefore, is 28 m2.
system turned off. Using the Boner concept of From the previous chapter, we know how to
system delta, the situation at maximum gain calculate the critical distance for a person talking
corresponds to a delta of unity. (Delta is defined as (nominal directivity index of 3 dB). In the example
the difference in decibels between sound level at the given, DC for a source having a directivity index of 3
system microphone with system off and the level in dB is 1 meter.
the audience area with system on. See Figure 6-1). The figure also shows geometrical relationships
Although we have described these as among a talker, a listener, the talker’s microphone
conditions of maximum potential system gain, it is and a simple wall-mounted loudspeaker having a
possible in practice to achieve a delta greater than directivity index of 6 dB along the axis pointed at the
unity. For example, if a directional microphone is listener. The microphone is assumed to be
used it can discriminate against the reverberant field omnidirectional.
6-2
Sound System Design Reference Manual
Step 1: Calculate relative sound levels produced meter, it must be 4.6 dB less at a distance of 1 meter,
by the talker at microphone and listener. or 65.4 dB, and the level of the reverberant field must
We begin with the sound system off. Although also be 65.4 dB. The sound level produced by the
the calculations can be performed using only relative unaided talker, at the listener’s position, therefore is
levels, we will insert typical numbers to get a better 65.4 dB.
feel for the process involved. At this point, let us consider two things about
The microphone is .6 meter from the talker, and the process we are using. First, the definition of
at this distance, the direct sound produces a level of critical distance implies that sound level is to be
about 70 dB. Since DC for the unaided talker is only measured with a random-incidence microphone. (For
1 meter, the microphone distance of .6 meter lies in example, we have chosen a non-directional system
the transition zone between the direct field and the microphone so that it indeed will “hear” the same
reverberant field of the talker. By referring to Figure sound field as that indicated by our calculations).
6-3, we note that the combined sound levels of the Second, we have worked with fractions of decibels to
reverberant field and the direct field at a distance of avoid confusion, but it is important to remember that
.6 meter must be about 1 dB greater than the direct the confidence limits of our equations do not extend
field alone. Therefore, since we have assumed a beyond whole decibel values, and that we must
level of 70 dB for the direct field only, the total sound round off the answer at the end of our calculations.
level at the microphone must be 71 dB.
Next, we use a similar procedure to calculate Step 2:The sound field produced by the
the sound level at the listener’s position produced by loudspeaker alone.
an unaided talker: Now let us go back to our example and
The listener is located 4.2 meters from the calculate the sound field produced by the
talker, more than 3 times the critical distance of 1 loudspeaker. Our system microphone is still turned
meter, and therefore, well into the reverberant field of off and we are using an imaginary test signal for the
the talker. We know that the sound level anywhere in calculations. We can save time by assuming that the
the reverberant field is equal to that produced by the test signal produces a sound level at the microphone
direct field alone at the critical distance. If the level of 71 dB — the same previously assumed for the
produced by direct sound is 70 dB at a distance of .6 unaided talker.
6-3
Sound System Design Reference Manual
The loudspeaker is mounted at the intersection acoustic gain of the system for this specific set of
of wall and ceiling. Its directivity index, therefore, is conditions must be the difference between the two,
assumed to be 6 dB. In this room, the critical or only 2.6 dB. Obviously such a sound
distance for the loudspeaker is 1.4 meters. This is reinforcement system is not worth turning on in the
almost the same as the distance from the first place.
loudspeaker to the microphone. Since the Note that system acoustical gain is dependent
microphone is located at the loudspeaker’s critical upon the distance from the microphone to the talker.
distance, and since we have assumed a level of 71 A more general concept is that of system delta.
dB for the total sound field at this point, the direct According to the Boner paper, the maximum
field at the microphone must equal 71 dB minus 3 theoretical D of a properly equalized system is unity.
dB, or 68 dB. In our example, D works out to be -3 dB. Why?
The listener is 4.8 meters from the loudspeaker The Boners emphasize that for maximum
(more than 3 times the critical distance) and system gain the microphone must be in the direct
therefore, well into the reverberant field of the field of the talker and in the reverberant field of the
loudspeaker. We know that the level in the loudspeaker. But in our example, the microphone is
reverberant field must equal the level of the direct not quite in the direct field of the talker and is located
field alone at the critical distance. The sound level at at the critical distance of the loudspeaker! To achieve
the listener’s position produced by the loudspeaker more gain, we might move the microphone to a
must, therefore, be 68 dB. distance .3 meter from the talker and use a more
directional loudspeaker. This would result in a 3 dB
Step 3: Potential acoustic gain is now considered. increase in D and a potential acoustic gain at the
Since we deliberately set up the example to listener’s position of about 9 dB.
represent the condition of maximum theoretical gain In practice, however, we cannot operate the
for a properly equalized system, we can use these system at a point just below sustained feedback.
same figures to calculate the difference in level at the Even if we modify the system as described above,
listener’s position between the unaided talker and the our practical working gain will only be about 3 dB.
talker operating with the system turned on. We have Our calculations merely prove what we could have
calculated that the unaided talker produces a level at guessed in advance: in a room this small, where an
the listener’s position of 65.4 dB. We have also unaided talker can easily produce a level of 65 dB
calculated that the level produced by the throughout the room, a sound system is unnecessary
loudspeaker at the listener’s position is 68 dB. The and of no practical benefit.
Figure 6-3. Relative SPL vs. distance from source in relation to critical distance
6-4
Sound System Design Reference Manual
Calculations for a Medium-Size Room the figure given by the equation R = Sa/(1 -a).
Rather than complicate the example, however,
Consider a more typical (and more assume that the equation really does work and that
complicated) situation in which the sound system is the room constant is about 110 m2.
used in a larger room and in which a directional The next step is to calculate critical distances
microphone is employed. Figures 6-4 and 6-5 show a for the talker and the loudspeaker. Since the
room having a volume of 918 m3, a total surface area loudspeaker does not have a uniform radiation
of 630 m2 and a = 0.15. pattern, we must calculate its critical distance at the
The first step is to calculate the room constant, particular angle in which we are interested. Figure
and we would do well to examine the actual 6-5 shows the distances involved and the
distribution of absorptive material in the room. geometrical relationships between talker,
Chapter 5 explains why the effective room constant microphone, loudspeaker and listener.
R’ in a particular situation may vary substantially from
6-5
Sound System Design Reference Manual
In the frequency range of interest, the The third step is to make similar calculations for
loudspeaker is assumed to have a directivity index the loudspeaker alone. The listener is located on the
along its primary axis of 9 dB. From Figure 6-6 we major axis of the loudspeaker and is more than 3
find the corresponding critical distance of 4.2 meters. times the critical distance of 4.2 meters. The
The loudspeaker’s directivity index at a vertical angle microphone is located at a vertical angle of 60
of 60° is assumed to be -3 dB, with a corresponding degrees from the loudspeaker’s major axis, and also
critical distance of 1 meter. The unaided talker has a is more than 3 times the critical distance (at this
directivity index of 3 dB and his critical distance must angle) of 1 meter. Both the listener and the
therefore be 2 meters. microphone are located in the reverberant field of the
Our next step in calculating system gain is to loudspeaker.
find the difference in level produced by an unaided If the sound level produced by the loudspeaker
talker at the listener position as contrasted with that at the microphone can be no greater than 70 dB (the
at the microphone position. In this example the same level as the talker) then the level produced by
listener is 12 meters from the talker and the the loudspeaker at the listener’s position must also
microphone again is .6 meters away. be 70 dB, since both are in the reverberant field.
The talker’s critical distance of 2 meters is more Having established these relationships we
than 3 times the microphone distance. Therefore, the know that the talker produces a level at the listener’s
microphone is well in the direct field of the talker. The position of 60 dB with the sound system off and 70
listener is more than 3 times the critical distance and dB with the sound system on, or a maximum
is well into the reverberant field of the unaided talker. potential gain of 10 dB. Allowing 6 dB headroom in a
Setting the level produced by the unaided talker at properly equalized system, we still realize 4 dB gain
70 dB for a distance of .6 meters, we calculate that at the listener’s position, and the sound system can
the direct field at DC must be 60 dB, and since the be said to provide a small but perceptible increase in
reverberant field must also equal 60 dB, the level sound level.
produced by the unaided talker at the listener’s
position is 60 dB.
6-6
Sound System Design Reference Manual
However, all of the preceding calculations have us that even though the loudspeaker is 75° off the
assumed that the microphone is an omnidirectional major axis of the microphone, it still provides 3 dB of
unit. What happens if we substitute a directional discrimination in favor of the direct sound from the
microphone? Figure 6-7 shows the additional loudspeaker.
geometrical relationships needed to calculate the We know that the loudspeaker’s directivity
increase in gain produced by a directional index is -3 dB along the axis between the
microphone. loudspeaker and the microphone. We also know that
Note that the distance from talker to the microphone’s directivity index along this axis is
microphone is still .6 meters and that the talker is +3 dB. The combined directivity indices along this
assumed to be located along the major axis of the axis must therefore, be 0 dB and we can find the
microphone. The loudspeaker is located 5.4 meters equivalent critical distance from Figure 6-6.
from the microphone along an angle of 75° from the The combined critical distance of loudspeaker
major axis. and microphone along their common axis is about
Figure 6-7 also shows a typical cardioid pattern 1.3 meters. Since the distance between the two is
for a directional microphone. The directivity index of more than 3 times this figure, the microphone still lies
such a microphone along its major axis is about 5 dB. within the reverberant field of the loudspeaker. Using
Since the talker is located on the major axis of the directional microphone should therefore allow an
the microphone, it “hears” his signal 5 dB louder than increase in potential system gain before feedback of
the random incidence reverberant field. In theory this about 5 dB. (In practice, little more than 3 or 4 dB of
should increase potential system gain by a factor of additional gain can be achieved.)
5 dB.
But we must also consider the microphone’s
directional characteristics with relation to the
loudspeaker. If the directivity index of the microphone
at 0° is 5 dB, the polar pattern indicates that its
directivity index at 75° must be about 3 dB. This tells
6-7
Sound System Design Reference Manual
6-8
Sound System Design Reference Manual
From our calculations of critical distances, we Below 500 Hz the response of the system can
see that the unaided talker will produce a sound level be gradually shelved, or attenuated, without seriously
at the listener of 59 dB in an empty room and about degrading the quality of speech. Above 4 kHz sound
55 dB with a full audience. For a usable working systems tend to take care of themselves, due to the
delta of -6 dB, the calculated acoustic gain at the increase in overall acoustical sound absorption. At
listener’s position is about 5 dB in an empty room very high frequencies, most environments are
and about 9 dB when full. substantially absorptive, the air itself contributes
Can we get more gain by turning off the considerable acoustical absorption and loudspeaker
loudspeaker directly over the microphone? Not in a systems tend to become directional. These factors
densely packed array such as this. The loudspeakers make it highly unusual to encounter feedback
are mounted close together to produce a uniform frequencies much above 2500 Hz.
sound field at ear level. As a result, the contribution To make sure that a sound reinforcement
of any one loudspeaker is relatively small. However, system will successfully amplify speech, it is a good
by turning off all the loudspeakers in the performing idea to make gain calculations in at least two
area and covering only the audience, some increase frequency bands. In a well-designed system, if
in system gain may be realized. calculations are made for the regions centered at 1
In the example just given, each loudspeaker is kHz and 4 kHz, chances are that no unforeseen
assumed to have a directivity index in the speech problems in achieving desired system gain will be
frequency region of +6 dB at 0°, +3 dB at 45°, and encountered.
0 dB at 60°. Suppose we use only the 25 However, the region below 500 Hz cannot
loudspeakers over the audience and turn off the 15 simply be ignored. The room constant and the
loudspeakers in the front of the room. In theory, the directivities of the loudspeaker system and the
increase in potential gain is only 1 dB with a single microphone should be checked in the 200 - 500 Hz
listener or 2 dB when the audience area is filled. range to make sure that there are not substantial
Even if we allow for the probability that most of the deviations from the calculations made at 1 and 4
direct sound will be absorbed by the audience, it is kHz. If the room has very little absorption below 1
unlikely that the gain increase will be more than 3 dB. kHz, and if the loudspeaker system becomes
The calculations required to arrive at these nondirectional in this region, it may be impossible to
conclusions are tedious but not difficult. The relative achieve satisfactory system gain without severely
direct sound contribution from each of the attenuating the mid-bass region. The result is the all
loudspeakers at microphone and listener locations is too familiar system which provides satisfactory
calculated from knowledge of polar patterns and speech intelligibility, but which sounds like an
distances. By setting an arbitrary acoustic output per amplified telephone.
loudspeaker, it is then possible to estimate the sound
level produced throughout the room by generally
reflected sound (reverberant field) and that produced The Indoor Gain Equation
by reflected plus quasi-direct sound.
From the foregoing discussions, we can
appreciate the complexity of indoor system gain
System Gain vs. Frequency Response analysis and the need for accurately calculating the
attenuation of sound along a given path, from either
In the preceding examples we have not defined talker or loudspeaker, noting when we leave the
the frequency range in which gain calculations are to direct field and make the transition into the
be made. In most sound systems the main reason for reverberant field. If we were to attempt to establish a
worrying about system gain is to make sure that the general system gain equation, we would have a very
voice of a person talking can be amplified sufficiently difficult task. However, in the special case where the
to reach a comfortable listening level in all parts of microphone is in the talker’s direct field, and both
the seating area. Therefore, the most important microphone and listener are in the loudspeaker’s
frequency band for calculating gain is that which reverberant field, then the system gain equation
contributes primarily to speech intelligibility: the simplifies considerably.
region between 500 and 4000 Hz. Let us consider such an indoor system, first
with the system turned off, as shown in Figure 6-10.
If the talker produces a level L at the microphone,
then the level produced at the listener will be:
6-9
Sound System Design Reference Manual
Level at listener = L - 20 log (Dct/Ds), where Dct Measuring Sound System Gain
is the critical distance of the talker. The assumption
made here is that the level at the listener is entirely Measuring the gain of a sound system in the
made up of the talker’s reverberant field and that that field is usually done over a single band of
level will be equal to the inverse square component frequencies. It is normally specified that system gain
at Dct. shall be measured over the octave-wide band
Now, the system is turned on, and the gain is centered at 1 kHz. Another common technique is to
advanced until the loudspeaker produces a level L at use pink noise which is then measured with the
the microphone. At the same time, the loudspeaker A-weighted scale. A typical specification for sound
will produce the same level L at the listener, since system gain might read as follows:
both microphone and listener are in the “The lectern microphone shall be used in its
loudspeaker’s reverberant field. normal position. A small loudspeaker shall be
Subtracting the levels at the listener between mounted on a stand to simulate a person talking
the system on and the system off, we have: approximately .6 meters from the microphone. The
response of this test loudspeaker shall be reasonably
Difference = L - [L - 20 log (Dct/Ds)] flat over the range from 250 - 4000 Hz.
“With the system turned off, the test
or: loudspeaker shall be driven with a pink noise signal
Gain = 20 log Dct - log Ds to produce a sound level of about 80 dB(A) at the
system microphone. This level shall be measured
Finally, adding a 6 dB safety factor: with a precision sound level meter, using the “A”
scale, with its microphone immediately adjacent to
Gain = 20 log Dct - 20 log Ds - 6 the sound system microphone.
“After noting the sound level at the system
Note that there is only one variable, Ds, in this microphone with the sound system turned off, the
equation; Dct is more or less fixed by the directivity of sound system shall be turned on and its gain advanced to
the talker and the acoustical properties of the room. a point just below sustained oscillation. The amplified
Of course there are many systems in which the sound level shall be measured with the same sound
microphone may be placed in the transition zone level meter in the central part of the auditorium.
between the talker’s direct and reverberant fields, or “The D of the sound system shall be calculated
where the listener is located in the transition region by subtracting the measured SPL at the microphone
between the loudspeaker’s direct and reverberant (system off) from the measured SPL in the auditorium
fields. In these more complicated cases, the (system on).”
foregoing equation does not apply, and the designer The gain of the system is of course measured at
must analyze the system, both on and off, pretty some point in the auditorium and is the level difference
much as we went stepwise through the three at that point produced by the test loudspeaker before
examples at the start of this chapter. and after the system has been turned on. Details of the
measurements are shown in Figure 6-11.
6-10
Sound System Design Reference Manual
6-11
Sound System Design Reference Manual
The last point is illustrated graphically in Figure ignored in this one simple calculation. The chart
6-12, adapted from the Peutz paper. Each of the assumes that satisfactory loudness can be achieved
diagonal lines corresponds to a particular and that there is no problem with interference from
reverberation time. Each shelves at a point ambient noise. It also postulates a single source of
corresponding to a direct-to-reverberant sound ratio sound and a well behaved, diffuse reverberant sound
of -10 dB. Note that the shelf may lie above or below field.
the 15% figure depending upon the reverberation The data from the Peutz paper have been
time of the room. This agrees with other published recharted in a form more convenient for the sound
information on intelligibility. For example, Rettinger contractor in Figure 6-13. Here we have arbitrarily
points out that in rooms having a reverberation time labeled the estimated intelligibility of a talker or a
of 1.25 seconds or less, direct sound and early sound system as “satisfactory”, “good”, or “excellent”,
reflections always make up the greater portion of the depending upon the calculated articulation loss of
total sound field. Intelligibility in such rooms is good consonants.
regardless of the direct-to-reverberant sound ratio at There often is a dramatic difference in the
any given listening position. Conversely, anyone who acoustical properties of a room depending upon the
has worked in extremely large reverberant spaces size of the audience. Calculations should be made
such as swimming pools or gymnasiums knows that on the basis of the “worst case” condition. In some
intelligibility deteriorates rapidly at any point much highly reverberant churches particularly, it may turn
beyond the critical distance. According to the chart, a out that there is no practical way to achieve good
15% articulation loss of consonants in a room having intelligibility through the entire seating area when the
a reverberation time of 5 seconds corresponds to a church is almost empty. The solution may involve
direct-to-reverberant sound ratio of only - 5.5 dB. acoustical treatment to lessen the difference between
Problems associated with speech intelligibility a full and an empty church, or it may involve a fairly
in enclosed spaces have received a great deal of sophisticated sound system design in which
attention prior to the publication of the Peutz paper. reinforced sound is delivered only to the forward
The virtue of Peutz’ method for estimating speech pews when the congregation is small (presuming that
intelligibility is its simplicity. It must be remembered, a small congregation can be coaxed into the forward
however, that a number of contributing factors are pews).
6-12
Sound System Design Reference Manual
Also, local acoustical conditions may exist A sound reinforcement system cannot be
which are not taken into account by statistical theory turned up indefinitely. In many situations it is difficult
and, therefore, not covered by the Peutz findings or enough to achieve a useful operating level of 60 - 65
any of the other equations we have studied. Such dB(A) without feedback. It is easy to see, therefore,
localized dead spots or zones of interference may that the presence of excessive background noise can
not be discovered until the sound system is installed. render an otherwise good sound reinforcement
In large reverberant spaces, sufficient flexibility system unsatisfactory.
should always be built into the sound system design As an example of how the Peutz analysis can
to allow for such surprises. dictate the type of sound system to be used, let us
The effect of masking by unwanted background consider a reinforcement system to be used in a large
noise has been touched on only briefly in this reverberant church. Details are shown in Figure 6-14.
section. Such unwanted noise may be produced by Let us assume that the reverberation time is 4
sound from the outside environment, by noisy air seconds at mid-frequencies and that the designer’s
handling equipment, by noisy backstage mechanical first choice is a single-point loudspeaker array to be
equipment or by the audience itself. For good placed high above the chancel. Coverage
listening conditions, the level of ambient noise as requirements pretty much dictate the directional
measured on the “A” scale should be at least 10 dB characteristics of the array, and let us assume that
below the desired signal. Since the optimum level for the array will consist of two JBL Bi-Radial horns: 20°
reproduced speech in the absence of strong by 40° for far coverage, and 90° by 40° for near
background noise is 65 - 70 dB(A) this means that coverage. What we wish to calculate is the direct-to-
background noise with a full audience should not reverberant ratio at selected points in the audience
exceed 55 dB(A). In auditoriums and concert halls, area to determine if the Peutz criteria for acceptable
acoustical designers normally attempt to reduce intelligibility can be met. The most direct way of doing
background noise in an empty house to a level not this is to calculate the total reverberant level in the
exceeding 25 dB(A). In a church or meeting hall, the room for a given power input to each horn and
maximum tolerable background noise for an empty compare it with the direct sound coverage provided
room is about 40 dB(A). by each horn over its coverage angle.
6-13
Sound System Design Reference Manual
The analysis shown in Figure 6-14 indicates But the question remains: What kind of system
that when each of the two horns is powered by one will work in this large resonant room? Clearly, a
watt, the reverberant field in the room (read directly distributed system is called for. In such a system, a
from Figure 5-21) is 94 dB-SPL. The direct field level number of lower-powered loudspeakers are placed
provided by each horn over its coverage angle is on columns on each side of the church, each
about 85 dB-SPL. This produces a direct-to- loudspeaker covering a distance of perhaps no more
reverberant ratio of -9 dB, and an inspection of than 5 or 6 meters. In this way, the direct-to-
Figure 6-13 tells us that the system will have reverberant ratio can be kept high. If such a system
marginal intelligibility. Note that for 4 seconds of is further zoned into appropriate time delays, the
reverberation time, the direct-to-reverberant ratio effect will be quite natural, with subjective source
should be no less than about -7 dB if acceptable localization remaining toward the front of the listening
intelligibility is to be expected. This simple analysis space. Details of this are shown in Figure 6-15.
has told us that, on paper, we have designed a Again, we calculate the total reverberant level
sound system which will likely fail to satisfy the and compare it with the longest throw each
customer. loudspeaker will be called upon to handle. There are
Had the system consisted of a single horn, 14 loudspeakers, 7 on each side. Let us assume that
knowledge of its on-axis DI and Q could have led the efficiency of these loudspeakers is 1.2% and that
quickly to a determination of critical distance, and the their sensitivity is 95 dB, 1 watt at 1 meter. Feeding
direct-to-reverberant ratio could have been scaled one watt into each loudspeaker results in a total
from DC. However, for the composite array analyzed acoustical power of 14 x .012, or 0.17 watt. Again
here, there is no single value of DI or Q which can be using Figure 5-21, we observe that the reverberant
used, and a direct calculation of the overall level will be 92 dB-SPL. The longest throw each
reverberant level, using what we know about the loudspeaker has to cover is, say, 4 meters. Since the
efficiency of the transducers, and making a 1-watt, 1-meter sensitivity is 95 dB, the direct field for
comparison with the direct field, based on the each loudspeaker will be 12 dB lower, or 83 dB.
sensitivities of the transducers, is the quickest way to
solve the problem.
6-14
Sound System Design Reference Manual
Thus, the direct-to-reverberant ratio will be We can also ask the question of whether our
83-92, or -9 dB. This is still not good enough, but we analysis using R’ would have materially affected the
must remember that more than half the listeners will performance of the central array system. A rigorous
be closer to a loudspeaker than 4 meters. Another analysis would be a little tedious, but we can make a
very important point we have not yet considered is simplifying assumption. Let us assume that half of
the fact that the distributed loudspeakers are aimed the direct sound from the central array was incident
almost totally into the audience, with its absorption on the audience with its .95 absorption coefficient.
coefficient considerably greater than a of .12. This is Let us round this off and call it 1.0 instead, resulting
the appropriate time to use R’ instead of R in our in no sound at all being reflected from the audience.
calculations. This would only lower the reverberant level in the
Calculating R’ based upon an a’ of .95 for the room by 3 dB, hardly enough to make the direct-to-
audience area in the 1 kHz band: reverberant ratio workable.
More than any other we have carried out in this
R’ = Sa/(1 - a’) = 375/.05 = 7500 m2. chapter, this analysis points up the multi-dimensional
complexity of sound system design. Again, we state
Recalculating the reverberant level from Figure that there are no easy solutions or simple equations.
5-21, we get 80 dB-SPL. The new direct-to- Instead, there is only informed rational analysis and
reverberant ratio is 83 - 80, or +3 dB, and the system thoughtful balancing of many factors.
will be quite workable.
Will the reverberant level really be only 80 dB?
In actuality, we might observe something a little
higher than 80 dB, but not enough to alter our
analysis significantly.
6-15
Sound System Design Reference Manual
The Role of Time Delay in Sound Figure 6-16E shows how delay is typically
Reinforcement implemented in sound reinforcement. Here, that
portion of the audience seated under the balcony
The preceding example mentioned time delay does not get adequate coverage from the central
as a means of preserving naturalness in a distributed array. Small loudspeakers placed in the balcony soffit
system. This comes about by way of the Haas (or can provide proper coverage only if they are delayed
precedence) effect (5), which is illustrated in Figure so that the sound arrives at the listeners in step with
6-16. If two loudspeakers are fed the same signal, a that from the central array. In this way, the listener
listener mid-way between them will localize the tends to localize the source of sound at the central
source of sound directly ahead (A). At B, we have array — not at the soffit loudspeakers. If the soffit
introduced a delay in one of the otherwise identical loudspeakers are not delayed, listeners under the
channels, and the listener will clearly localize toward balcony would localize sound directly overhead, and
the earlier loudspeaker. At C. the leading signal has those listeners just in front of the balcony would be
been reduced in level, resulting in an effect of equal disturbed by the undelayed sound. In practice, the
loudness at both loudspeakers. This has the delay is usually set for an additional 20 msec in order
approximate effect of restoring the apparent to minimize comb filtering in the overlap zone
localization to the center. While this tradeoff is not an between direct and delayed sound fields.
exact one, the values shown in the graph at D The ready availability of solid state digital delay
indicate the approximate trading value between level units has made time delay an indispensable element
and delay for equal loudness at both loudspeakers. in sound system design.
6-16
Sound System Design Reference Manual
Figure6-18.Systemequalization
6-17
Sound System Design Reference Manual
At A, we see the on-axis (solid curve) and This spread between reverberation times at
power (dotted curve) response of a 2-way system low, mid, and high frequencies is typical of a good
making use of a ported LF horn unit and an older auditorium. When we calculate the room constant as
type HF radial horn. When such a system is a function of frequency and plot it, along with the
equalized for smooth power response, as in the case sound level that would be produced by one acoustic
of the standard mid-house equalization procedure, watt in the room, we see that the total variation in
then the on-axis, or direct field response of the SPL is only about 3 dB. The importance of this
system will have a couple of “bumps” in its response. observation is that, if we had a loudspeaker system
This will have the effect of making both speech and exhibiting flat power response, then it would produce
music sound unnatural. a reverberant SPL in this auditorium that would vary
Now let us examine the case at B. Here, the LF no more than the inverse of the curve shown in
part of the system consists of a single 380 mm (15”) Figure 6-19. Obviously, the smoother the power
LF driver in a ported enclosure, and the HF horn is a response of a loudspeaker, the less equalization it
JBL 2360 Bi-Radial. Note that the power response will require and the more natural it will sound on all
and on-axis response very nearly lie over each other. types of program.
Thus, the adjustment of the system out in the house Another use of equalization is in controlling
will result in both reverberant field response (power feedback. As we have stated many times, a sound
response) and direct field response (on-axis reinforcement system should be operated at least 6
response) tracking each other closely. Such a system dB below the point of feedback if it is to be stable.
can often be broad-band-equalized merely through Through careful and selective use of narrow-band
the proper choice of components, dividing network notch filters, the first several ring modes of a sound
and transducer drive levels, requiring little, if any, system can be minimized, and the overall system
added electronic equalization. gain can be increased perhaps 3 or 4 dB. The
The graph shown in Figure 6-19 shows this practice of narrow-band equalization is complex, and
clearly. Here, we have plotted the variation in R over it is best left to those who have been trained in it.
the frequency range for a large auditorium. The room
we have chosen has the following characteristics:
V = 13,500 m3
S = 3538 m2
6-18
Sound System Design Reference Manual
6-19
Sound System Design Reference Manual
Sound System Design Reference Manual
Just as the building architect interprets a set of Assume that we have the following
requirements into flexible and efficient living or requirements:
working spaces, the designer of a sound 1. Up to ten microphones may be needed at
reinforcement system similarly interprets a set of different locations.
requirements, laying out all aspects of the system in 2. The system is to be used primarily for
an orderly fashion. A full sound system specification speech reinforcement.
will detail almost everything, including all equipment 3. The system shall be able to produce peak
choices and alternatives, rack space requirements, levels up to 85 dB-SPL in all parts of the house under
wire gauges and markings, and nominal signal all speech input conditions, including weak talkers.
operating levels. In addition, the electroacoustical The room noise level is about 25 dB(A).
aspects of the system will have been worked out well The most basic interpretation of these
ahead of time so that there will be few surprises requirements tells us the following:
when the system is turned on for the first time. 1. A small Soundcraft or Spirit console should
The consultant or design engineer lays out the suffice for all input configurations and routing control.
broad system parameters, but it is the sound 2. A single central array is the preferred system
contractor who is responsible for all component type, based on the desire for most natural speech
layout and orderly completion of the system, along reproduction. The array may be specified using
with documentation for usage as well as individual HF and LF components; alternatively, an
maintenance. System architecture also addresses appropriate full-range system with integral rigging
signal flow and nominal operating levels, consistent capability may be specified, as we will show here.
with the requirements of the system. The best 3. Both biamplification and system response
designs are usually the simplest and most equalization are recommended, and this suggests
straightforward ones. that a digital loudspeaker controller be used for
In this chapter we will cover several design frequency division, time alignment, and system
projects, beginning with basic design goals and response equalization.
fundamental performance specifications. We will then Note that there are many points in the system
move on to system descriptions and layout, where we can set or change gain. There is always
suggesting ways that the specification can be met. considerable gain overlap in the electronic devices
We will concentrate on the electroacoustical used in sound system work. The purpose of this is to
problems that are fundamental to each case study. allow for a great variety of input conditions as well as
By way of review, we will first discuss a few basic to allow the equipment to be configured in different
audio engineering subjects, beginning with an ways, as required. It is critical that the designer
abbreviated signal flow diagram for a relatively specify a nominal setting of each gain control,
simple speech reinforcement system. locking off, when possible, those controls that will not
— or should not — be altered during normal system
use. This important setting of gain relationships
should be based on the absolute requirement that
the input noise floor of the system should not be
degraded later in the chain, and that no early stage
of amplification should overload before the output
power amplifier overloads. In our exercise here, we
7-1
Sound System Design Reference Manual
will simplify things by considering only a single setting represents a nominal operating point for the
microphone path through the system to a single microphone/console combination, and there is ample
loudspeaker. flexibility for operating the system above or below
For the moment, let us consider only the this setting, as may be required by weak or loud
abbreviated console flow diagram shown in the talkers. Frequency division and system equalization
upper part of Figure 7-1A. Microphone ratings in use are to be carried out by a digital controller, the JBL
today state the unloaded output voltage when the model DSC260. The loudspeaker to be used is the
unit is placed in a sound field of 94 dB SPL. Normal JBL model SR4726A, and the recommended
speech level at an operating distance of .5 meter is amplifier is the JBL model MPX600. Typical
about 72 dB SPL; If we are using a microphone with operating levels are as shown in the lower portion of
a sensitivity of 10 mV/Pa, the microphone’s nominal Figure 7-1A.
voltage output in the 72 dB sound field will be: The level diagram shown in Figure 7-1B shows
that, at the power amplifier’s output, the noise level
E = 1022/20 x 10 mV = .8 mVrms of the microphone is about 3 dB greater than the
noise contributed by the power amplifier. Both of
Step One: these noise sources will be swamped out by the
Set a reference input of .8 mVrms at 1000 Hz at acoustical noise level in the acoustical space,
one of the microphone inputs on the console. With however. The electrical noise floor is transformed
the input and output faders at their nominal “zero” over to an equivalent noise level of -2 dB(A) at a
markings, set the microphone’s input trim control for distance of 20 meters, some 25 dB lower than the
a console output of 0.4 Vrms. (Alternatively, a stable acoustical noise floor of a typical space. With this
sound pressure level of 72 dB may be generated at calibration procedure, the maximum output level
the microphone, and the microphone trim setting possible in the house is limited by the dynamic range
adjusted for 0.4. Vrms output.) In making this setting, and nominal operating point established for the
the trim potentiometer marker will normally be DSC260. If more output level is desired, the nominal
somewhere between 10 o’clock and 2 o’clock. This operating points must be reset accordingly.
7-2
Sound System Design Reference Manual
Figure 7-1B. detailed level diagram showing noise levels, nominal operating levels,
and maximum output levels of each device
7-3
Sound System Design Reference Manual
Step Two: see that for an ambient noise level in the 25 dB(A)
We now have to determine what the nominal range, the EAD for a lowered voice would be about
operating level of the system should be for the 2 meters, or a speech level of about 60 dB SPL.
farthest listeners, which we will assume are some 20 For a direct field level of 60 dB at a distance of
meters away from the loudspeaker. Let us further 20 meters, the LF section of the loudspeaker will
assume that the reverberation time in the room is no require a signal input of 0.1 watts (into 8 ohms). In
greater than 1.5 seconds in the range from 250 Hz to the biamplification mode the HF section will require
2 kHz and that the average noise level room is in the considerably less than 0.1 watt input in order to
range of 25 dB(A). Referring to Figure 7-2, we can reach the desired level at a distance of 20 meters.
7-4
Sound System Design Reference Manual
7-5
Sound System Design Reference Manual
Constant Voltage Distribution Systems designer (or installer) merely has to keep a running
(70-volt lines) tally of watts drawn from the line, and when the
number of watts equals the continuous output power
Many distribution systems in the United States rating of the amplifier, then the system is fully loaded.
make use of the 70-volt line for powering multi- Ordinarily, no additional loads will be placed across
loudspeaker paging systems. In Europe the 100-volt the line, but there is some leeway here.
line is common. In either system, the full output The alternative to 70-volt distribution is to
power of the driving amplifier is available at a line laboriously keep track of combined load impedances
voltage of 70 Vrms or 100 Vrms, respectively. in parallel, a big task. Details of a 70-volt transformer
In placing loads across the line, the design are shown in Figure 7-5.
engineer simply keeps a running count of the number In Europe, a 100-volt transmission system,
of watts of power drawn from the line. There is no derived in a similar manner, is used.
need to calculate the aggregate load impedance at
any point in the process. When the total number of
watts drawn from the line is equal to the power rating Low Frequency Augmentation —
of the amplifier, then the line is fully loaded and Subwoofers
properly matched.
Figure 7-4 shows details of a 70-volt distribution Whether in the cinema or in open spaces, LF
system. The maximum load on the amplifier is augmentation systems are becoming popular for
transformed so that the applied voltage will be 70 special effects. For indoor applications many
Vrms. This then will correspond to a total transformed acoustical engineers calculate the reverberant sound
load impedance, ZL, equal to 5000/PO, where PO is pressure level that can be produced by a transducer,
the maximum power output of the amplifier. or group of transducers, operating continuously over
Individual loads are placed across the amplifier an assigned low frequency band, normally from 25
in parallel using line-to-loudspeaker distribution Hz to about 80 Hz. The equation for determining the
transformers that have a 70-volt primary and a reverberant level is:
tapped secondary designated in watts. The system
7-6
Sound System Design Reference Manual
LREV = 126 + 10 log WA - 10 log R, Some design engineers prefer to make actual
direct field calculations for one or more subwoofer
where WA is the continuous acoustical power output units at a distance, say, of two-thirds the length of the
from the transducer and R is the room constant in m2. enclosed space. In large motion picture spaces, both
In using this equation, we assume that the sets of assumptions yield results that are usually
space is fairly reverberant at very low frequencies within 5 dB of each other.
and that the value of absorption coefficient at 125 Hz The phenomenon of mutual coupling always
(the lowest value normally stated for materials) will comes to our aid in increasing the power output of
be adequate for our purposes. combined subwoofer units. Figure 7-6A shows the
7-7
Sound System Design Reference Manual
7-8
Sound System Design Reference Manual
Case Study A: A Speech and Music fascia of the balcony and the curved back wall itself.
System for a Large Evangelical Church: If not properly treated, these surfaces can cause
severe reflections back to the platform. In many
1. Basic Description and Specifications: cases, such spaces are designed with an
The fan shaped architectural design shown in overabundance of absorbing material, making the
Figure 7-7 is common for modern evangelical room quite dead. There is then a need for a very
churches in that it accommodates many people, all robust speech-music reinforcement system to
seating positions with good sightlines. The major provide a feeling of ambience and acoustical
acoustical problem is likely to be the curved front envelopment.
7-9
Sound System Design Reference Manual
7-10
Sound System Design Reference Manual
The other approach is to use pre-existing All units will have a nominal downward
building blocks, such as the SP225-9 system. elevation angle of about 60°, and the average throw
Specifically, four of these systems can be arrayed to the main floor is 10 meters. Level calculations are
with sufficient space between them for good stereo as follows:
presentation and splayed to produce an included
coverage angle of 135°. Rigging is integral in the SP Level Power Distance
Series, so that problem is solved. A center pair of 100 dB 1W 1 meter
SP225-9 units should be located side by side and 131 dB 1200 W 1m
splayed along their common back angle to give 120° 111 dB 1200 W 10 m
coverage. This channel is primarily for speech, but
may be used as well as a stereo center channel for In this case, we are powering the two LF units
music. The remaining two SP225-9 units will be used in each SP225-9 as a parallel (4 ohm) load to be
for left and right stereo presentation. driven by one section of a MPX1200 amplifier.
7-11
Sound System Design Reference Manual
The three HF sections in each main array Case Study B: A Distributed Speech
should be powered by one section of an MPX600 Reinforcement System for a Large
amplifier (200 watts into 16 ohms). Liturgical Church
Each one of the four main arrays should also
have an SP128S subwoofer module. These would 1. General Information and Basic Performance
be powered the same way as the LF sections of the specifications:
SP128S units. Level calculations are given: The system to be studied here is typical of what
may be found in just about every large religious
Level Power Distance edifice in Europe and in many large cities in the U. S.
102 dB 1W 1 meter The plan and front section views are shown in Figure
133 dB 1200 W 1m 7-10. The building under consideration here has an
113 dB 1200 W 10 m internal volume of 12,000 cubic meters and surface
area of 4000 square meters. A mid-band empty
The electrical diagram for the main array is house reverberation time of 2.5 seconds indicates a
shown in Figure 7-8. value of 800 square meters of absorption units (Sa),
as extrapolated in Figure 5-10.
3.2 Delay Rings: Our major concerns with a system in such a
The first delay ring of loudspeakers should be space as this are the net speech direct-to-
composed of fairly robust loudspeakers, such as the reverberant ratio and the reverberation time itself.
8340A. Each one of these eight loudspeakers has a Each loudspeaker will contribute to the overall
far throw range of about 6 meters and can reverberant level behind the amplified speech, and
accommodate 150 watts input. We can calculate the our first step is to determine the number of
level output of each one as follows: loudspeakers that will be required to cover the entire
seating area. Studying the plan view of the building,
Level Power Distance we can see that 8 loudspeakers will cover the
96 dB 1W 1 meter transept seating, while 10 systems will cover the
118 dB 150 W 1m nave seating.
102 dB 150 W 6m
2. Analysis:
As an ensemble, these eight loudspeakers will The longest “throw” that will be required of any
provide added high frequency coverage of the single loudspeaker is to cover a listener seated at
balcony seats, ensuring good articulation. the center aisle, a distance of about 7 meters. Let
The under balcony soffit system consists of 16 us now specify a JBL Control 28 and power it to
transducers located 2 meters above the seated produce a level of 85 dB at a distance on-axis of 7
listener’s ear height. The model 2142 has a meters.
sensitivity of 97 dB and a power rating of 90 watts. We can do this directly by setting up the familiar
A seated listener directly under one of these level/power/distance chart as follows:
transducers, if it is powered by 1 watt, will hear a
level of 91 dB. A listener mid-way between a pair of Level Power Distance
them will hear a level of about 90 dB. The 90-watt 92 dB 1W 1 meter
per transducer rating means that peak levels of 75 dB 1W 7m
about 110 dB can be developed under the balcony. 85 dB 10 W 7m
7-12
Sound System Design Reference Manual
Figure 7-10. Plan and front elevation views of large liturgical church
7-13
Sound System Design Reference Manual
Entering values and rearranging: For the next step in the analysis we need to
determine the resulting reverberant level in the room.
10 log Efficiency = 92 -109 - 5 = -22
Lrev = 126 + 10 log WA - 10 log R
The efficiency is then 10-22/10 = 10-2.2 = .63%
In relatively live spaces, Sa and R are virtually
The total contribution to the reverberant field identical; therefore,
will be from all 18 loudspeakers working at once. We
can then calculate the radiated acoustical power as: Lrev = 126 - 0.6 - 29 = 96 dB SPL.
18 x 10 x 0.0063 = 1.134 acoustical watts. Here, 18
is the number of individual loudspeakers, 10 W is the We have now reached a point in our analysis
electrical power applied to each of them, and 0.0063 where we can estimate the overall system
is the efficiency. performance regarding speech intelligibility. We know
the following:
7-14
Sound System Design Reference Manual
7-15
Sound System Design Reference Manual
Case Study C: Specifications for a 1.3 Social Area: This space is of irregular
Distributed Sound System Comprising a shape, as shown in the diagram. A foreground stereo
Ballroom, Small Meeting Space, and music system is to be specified for this space; no
Social/Bar Area. paging will be required. The system should be
capable of producing levels of 85 dB SPL. There is
1. General Information and Basic Performance also a disco/dance floor area, and a four-
Specifications: loudspeaker installation should provide levels of
1.1 Ballroom Description: The size of the space 105 dB at the center of the dance floor.
is 33 meters long, 22 meters wide, and 8 meters
high. A stage is located at the center of one short 2. Exercises:
side, and the room may be used for banquets, Study the attached figures that detail the layout
displays, and social events such as dancing. of distributed systems in general, and pick either the
A distributed (ceiling) system will be used for square or hexagonal layout.
general speech/music purposes, as well as
amplification of stage events. For this purpose the 2.1 Ballroom System:
system should be zoned for delay. Reinforced levels 1. Determine quantity and placement of ceiling
up to 100 dB SPL will be expected, and coverage loudspeakers that will meet the specification.
should be uniform within 1.5 dB up to a frequency of 2. Determine the power allocation for each
2 kHz. The space is normally carpeted, except for loudspeaker and describe the power distribution
dancing. Reverberation time is minimal. system (70-volt or low-Z).
1.2 Meeting Space Description: This space is 3. Determine the minimum number of workable
typical of many that will be found in convention and zones for signal delay for stage events.
meeting areas. The size is 8 meters by 5 meters and
3 meters high. A distributed ceiling system is to be 2.2 Meeting Space System:
designed, uniform within 1.5 dB up to 2 kHz. Normal 1. Determine the model loudspeaker required
maximum levels are expected to be 85 dB SPL. and the spacing density in the ceiling.
Figure 7-13. Ballroom layout. Plan view (A); side section view (B).
7-16
Sound System Design Reference Manual
7-17
Sound System Design Reference Manual
Figure 7-15. Meeting space layout. Plan view (A); side section view (B)
7-18
Sound System Design Reference Manual
9 8 89 - 18 = 71
>74.5
10 7 89 - 17 = 72
>77.5 >77.5
11 8 89 - 18 = 71
>74.5
12 7 89 - 17 = 72
7-19
Sound System Design Reference Manual
The total level at the test position is thus 82 dB each corner to the center of the dance floor is 7
when all 12 loudspeakers are powered with 1 watt, or meters:
85 dB with 2 watts per loudspeaker. With 20 watts
per loudspeaker, we would have a comfortable 10 dB Level Power Distance
margin over our target value of 85 dB SPL. The 98 dB 1W 1 meter
system will be “coasting” most of the time, and a 126 dB 600 W 1m
single stereo amplifier, with loads wired in series- 109 dB 600 W 7m
parallel, will suffice.
2. Disco system. There are a number of A total of 8 4892 systems would increase the
possibilities here. A very high-end approach would be output by 9 dB, producing level capability of 118 dB
to specify two JBL 4892 Array Series systems at at the center of the dance floor. The amplifier
each corner. They should be mounted near the requirement would be 4 MPX600 units and 4
ceiling and angled down so that the horns point MPX300 units. Note carefully that the MPX600 has a
toward the center of the dance floor. maximum output capability of 400 watts into 8 ohms.
The 4892 is a biamplified system, and its This is approximately 2 dB less than 600 watts, and
overall sensitivity of 98 dB is limited by its LF section. we would have to derate the system’s overall output
The LF power rating is 600 watts. The distance from capability by that amount.
7-20
Sound System Design Reference Manual
7-21
Sound System Design Reference Manual
7-22
Sound System Design Reference Manual
Recommended Reading:
1. D. and C. Davis, Sound System Engineering, second edition, Howard F. Sams, Indianapolis, 1987.
2. J. Eargle, Electroacoustical Reference Data, Van Nostrand Reinhold, New York, 1994.
3. Various, Sound Reinforcement, an anthology of articles on sound reinforcement from the Journal of the
Audio Engineering Society, volumes 1 through 26. (Available from the AES)
Papers:
4. C. P. and R. E. Boner, “The Gain of a Sound System,” J. Audio Engineering Society, volume 17, number 2
(1969).
5. H. Haas, “The Influence of a Single Echo on the Audibility of Speech,” J. Audio Engineering Society,
volume 20, number 2 (1972).
6. H. F. Hopkins and N. R. Stryker, “A Proposed Loudness-Efficienty Rating for Loudspeakers and the
Determination of System Power Requirements for Enclosures,” Proceedings of the IRE, March 1948.
7. C. T. Molloy, “Calculation of the Directivity Index for Various Types of Radiators,” J. Acoustical Society of
America, volume 20, pages 387-405 (1948).
9. V. M. A. Peutz, “Quasi-steady-state and Decaying Sound Fields,” Ingenieursblad, volume 42, number 18
(1973, in Dutch).
10. Various, “Loudspeaker Arrays — Design and Performance,” J. Audio Engineering Society, volume 38,
number 4 (1990).
© 1999 JBL Professional
JBLProfessional Printed in U.S.A.
8500 Balboa Blvd. SND DES MAN99
Northridge, CA 91329 U.S.A. rev. B
www .jblpro.com